The Ser/Thr Protein Kinase PknB Is Essential for Sustaining Mycobacterial Growth

Bacterial strains and growth conditions.Escherichia coli XL1-blue, used for cloning experiments, was routinely propagated in LB broth (Difco) at 37°C. M. smegmatis mc²155 (41) was grown at 30, 37, or 42°C in LB broth supplemented with 0.05% Tween 80 or in 7H9 broth (Difco) supplemented with 0.05% Tween 80 and albumin-dextrose-catalase (ADC). M. tuberculosis H37Rv was grown in 7H9 broth supplemented with 0.05% Tween 80 and ADC or on 7H11 agar medium supplemented with oleic acid-albumin-dextrose-catalase (OADC) at 30, 37, or 39°C. Antibiotics were added at the following concentrations: ampicillin, 100 μg/ml; kanamycin, 20 μg/ml; hygromycin, 50 μg/ml. When required, 10% (M. smegmatis) or 2% (M. tuberculosis) sucrose was added to the solid medium.

Cloning procedures and Southern blot analysis. (i) Construction of pRBexint-pknB and pOMK-pknB.Preparation of E. coli and mycobacterial electrocompetent cells, extraction of chromosomal DNA, Southern blotting, and cloning were carried out as described previously (24, 35). M. smegmatis genomic sequences were obtained from TIGR (http://www.tigr.org).

M. tuberculosis pknB and flanking regions were PCR amplified and cloned into the SpeI restriction site of the replicative plasmid pOMK (22) to obtain pOMK-pknB or were amplified using primers harboring SpeI and HpaI restriction sites and cloned into pRBexint (a kind gift from R. Brosch), an integrative cosmid vector derived from pYUB412 (5), to generate pRBexint-pknB. All constructs were verified by sequencing. Details of the primers are available upon request.

(ii) Construction of the M. tuberculosis and M. smegmatis pknB mutant.The essentiality of the pknB gene in M. tuberculosis and M. smegmatis was investigated following a standard strategy based on the construction of merodiploid strains (23). The ts-sacB method was used to achieve allelic replacement at the pknB locus of M. tuberculosis (23, 35), and pRBexint-pknB was integrated to perform complementation experiments. A unique HindIII restriction site was created in the M. tuberculosis pknB gene by using a QuikChange site-directed mutagenesis kit (Stratagene) into which the kanamycin resistance cassette from pUC4K (Amersham Biosciences) was inserted. The resulting pknB::Km gene was then cloned into pPR27 with the xylE colored marker (35) to obtain pPR27pknB, the construct used for allelic replacement.

The essentiality of the M. smegmatis pknB gene was investigated using a two-step homologous recombination procedure to achieve allelic replacement at the pknB locus (23) and pRBexint-pknB to perform complementation experiments. The M. smegmatis pknB gene and flanking regions were PCR amplified, and a disrupted allele, pknB::Km, was then obtained as described above. The construct used for allelic replacement, pJQpknB, was obtained by cloning pknB::Km into pJQ200-XylE (23).

Protein kinase assays.PknB and GarA were expressed in E. coli and purified as described previously (9, 43). Kinase assays were carried out in 15 μl of kinase buffer (50 mM HEPES [pH 7.0], 1 mM dithiothreitol, 0.01% Brij35, 5% glycerol, 2 mM MnCl₂). All reactions were started with the addition of ATP and conducted at 30°C for 20 min. Myelin basic protein (MBP) phosphorylation assays were carried out with an enzyme:substrate molar ratio of 1:20 in the presence of ATP (0.1 mM containing 1 μCi of [γ-³³P]ATP). The reactions were stopped by adding EDTA (15 mM, final concentration), and the proteins were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. For GarA phosphorylation, a 2.25 μM final concentration of ATP (containing 1 μCi of [γ-³³P]ATP) and an enzyme:substrate ratio of 1:2,000 were used. The reactions were stopped by heat inactivation of the enzyme, and 5 μl of the reaction was transferred by pipette onto P81 paper (phosphocellulose; Whatman). The P81 papers were washed with 1% phosphoric acid three times for 5 min, rinsed with acetone, and allowed to dry on a sheet of aluminum foil. The radiolabeled spots of both one-dimensional gels and P81 papers were visualized by autoradiography, and the incorporation of radiolabeled ATP was quantified using a PhosphorImager system (Storm; Molecular Dynamics). In inhibition experiments, each compound (100 μM in MBP phosphorylation assays; 10 μM in GarA phosphorylation assays) was preincubated for 30 min at 4°C in the reaction mixture without ATP. All compounds tested (A-3, calphostin C, GF109203X, H-7, H-8, H-9, H-89, HA-1004, HA-1077, hypericin, K-252-a, K-252-b, KN-62, KT-5720, KT-5823, ML-7, ML-9, and staurosporine) were purchased from BIOMOL Research Laboratories, Inc. For determination of the 50% inhibitory dose (IC₅₀), serial twofold dilutions were started at 10 μM (K-252-a and K-252-b) or 20 μM (staurosporine), and the IC₅₀ values and standard errors were calculated using KaleidaGraph (Synergy Software).

Resazurin microtiter assays.Resazurin sodium salt powder (Sigma) was prepared at 0.01% (wt/vol) in distilled water and filter sterilized. The assays were performed at 37°C in 7H9 medium containing 0.2% glycerol and OADC. Serial twofold dilutions of each inhibitor were prepared in 100 μl of medium directly in 96-well plates at concentrations of 40 to 0.31 μM for K-252-a and K-252-b and 200 to 1.56 μM for staurosporine. Growth controls containing no inhibitor and sterility controls without inoculation were also included. The inoculum was prepared from fresh cultures adjusted to an optical density at 600 nm of 0.1 in the medium described above and diluted 1:10, and 100 μl was added per well (∼10⁶ cells). The plates were incubated at 37°C in a 5% CO₂ incubator for 10 days (for M. tuberculosis) or 4 days (for M. smegmatis). Thirty microliters of resazurin solution was then added to control wells containing no inhibitor or medium alone, incubated overnight at 37°C, and assessed for color development. A change from blue to pink indicates the reduction of resazurin and, therefore, bacterial growth. If a color change occurred, resazurin was added to all of the wells. The MIC was defined as the lowest drug concentration that prevented the color change.

Genomics and experimental rationale.An inspection of the available genome sequences of mycobacteria, corynebacteria, and streptomycetes (1, 8, 11, 12, 20) revealed a conserved operon, near the chromosomal origin of replication, comprising five or six genes that may be involved in signal transduction pathways and cell division. In M. tuberculosis, this operon, in which the termination codon of each gene overlaps the initiation codon of its follower, begins with Rv0019, coding for forkhead-associated (FHA) protein B. This is followed by an in-frame mycobacterial intergenic repetitive unit (42) which precedes pstP, encoding phosphoserine/threonine protein phosphatase (9), and the rodA and pbpA genes before the operon ends with the STPK genes, pknA and pknB (Fig. 1). Except for the presence of the mycobacterial intergenic repetitive unit, which is confined to M. tuberculosis, the same gene arrangement occurs in all sequenced actinobacteria, and the operon is transcribed counter to the direction of replication. In Streptomyces coelicolor, there is one less STPK gene, as pknA appears to have been lost (7). It has been speculated that these STPK genes may be essential and control cell division (3), although there is no formal evidence. To substantiate this claim, we have undertaken genetic and biochemical analyses of pknB which, as the distal gene in the operon, should be readily inactivated, without polar effects, unless it is essential.

• Open in new tab
• Download powerpoint

FIG. 1.

Conserved structure of the fhaB-pknB gene cluster in actinobacteria. This cluster comprises six genes, except in Streptomyces coelicolor where pknA is missing: fhaA, encoding FHA protein A; fhaB, FHA protein B; pstP, phosphoserine/threonine protein phosphatase; rodA, cell division protein; pbpA, peptidoglycan biosynthesis protein; pknA-pknB, STPK. In all five bacteria, the cluster is situated near oriC but transcribed convergently with respect to the direction of replication. Abbreviations: M. tb., M. tuberculosis; M. smeg., M. smegmatis; C. diphth., Corynebacterium diphtheriae; S. coelicol., Streptomyces coelicolor.

Inactivation of pknB in M. tuberculosis.Attempts were made by allelic replacement to disrupt pknB in the pathogenic species, M. tuberculosis H37Rv, using the ts-sacB methodology (23, 35). However, despite our having performed three independent experiments, PCR analysis of clones corresponding to putative double crossover recombinants showed that none of them exhibited the expected pattern for allelic exchange mutants. To investigate whether the failure to disrupt the pknB gene could be due to its essentiality, we next performed an allelic replacement experiment using a M. tuberculosis merodiploid strain. This strain was constructed by integrating a functional copy of pknB into the chromosome of M. tuberculosis H37Rv using the integrative cosmid vector pRBexint-pknB. The integration of this cosmid at the tRNA^Gly site of the chromosome was confirmed by PCR and Southern blotting. Gene replacement experiments were carried out as described above. Analysis by PCR and Southern blotting of clones corresponding to putative double crossover recombinants showed that, in most cases, the allelic exchange occurred at the integrated pknB locus. However, we were also able to isolate clones in which the gene replacement had taken place at the wild-type pknB locus (Fig. 2a). Therefore, the expression of the wild-type pknB gene from pRBexint-pknB was sufficient to rescue a M. tuberculosis pknB knock-out mutant, indicating that this gene is essential for M. tuberculosis.

• Open in new tab
• Download powerpoint

FIG. 2.

(a) Allelic replacement at the M. tuberculosis pknB locus. Southern blot analysis and corresponding NheI restriction profiles of DNA from H37Rv (lane 1), DNA from the initial diploid strain carrying the pRBexint-pknB integrated plasmid (lane 2), and DNA from the resulting allelic replacement mutant strain (lane 3) are shown. The probe was a 1.9-kb PCR fragment carrying the pknB gene. (b) Allelic replacement at the M. smegmatis pknB locus. Southern blot analysis and expected hybridization profiles of a pknB allelic exchange mutant carrying the pRBexint-pknB rescue plasmid (lane 1) and mc²155 (lane 2) are shown. Chromosomal DNA was digested with PstI. The probe used to perform the hybridization corresponds to the wild-type pknB gene and flanking regions obtained by PCR (2.3-kb EcoRI/XbaI restriction fragment).

Inactivation of pknB in M. smegmatis.To investigate whether the pknB gene is also required for M. smegmatis growth, a kanamycin-disrupted copy of the pknB gene, pknB::Km, was inserted into the sacB suicide vector pJQ200-XylE and introduced into wild-type M. smegmatis mc²155 by electroporation. Kanamycin-resistant mc²pJQpknB transformants were selected with LB-kanamycin (LB-Km) plates at 30°C, and Southern blot analyses of eight of them indicated that all resulted from a single crossover event at the pknB locus (data not shown). Subsequently, two transformants, mc²pJQpknB.2 and mc²pJQpknB.5, were grown in LB-Km broth and then plated onto LB-Km-sucrose plates to select for clones that had undergone a second intrachromosomal crossover. Allelic exchange mutants are expected to carry the disrupted allele pknB::Km and to have lost the sacB and xylE genes carried by pJQpknB. However, the spraying of thousands of kanamycin-sucrose resistant colonies with catechol (to test for XylE⁻ strains) revealed that none of them exhibited the expected phenotype. Instead, these clones had probably undergone mutations in the sacB gene that conferred sucrose resistance.

To investigate whether the failure to disrupt pknB was due to its essentiality, we transformed the single crossover strains mc²pJQpknB.2 and mc²pJQpknB.5 with either pRBexint or pRBexint-pknB. Allelic exchange mutants were selected on LB-Km-hygromycin B-sucrose plates at 30°C as described above. The phenotype of approximately 1,000 Km^r-Hyg^r-Suc^r colonies was tested for both types of transformants plated. No allelic exchange mutant (Km^r-Hyg^r-Suc^r-XylE⁻ colony) was found when mc²pJQpknB.2/pRBexint or mc²pJQpknB.5/pRBexint was plated, confirming our previous results. In contrast, when mc²pJQpknB.2/pRBexint-pknB and mc²pJQpknB.5/pRBexint-pknB were plated, 30% of Km^r-Hyg^r-Suc^r colonies were found to be XylE negative. A Southern blot analysis of a recombinant revealed that it had undergone gene replacement at the pknB locus (Fig. 2b), indicating that pknB is also an essential gene in M. smegmatis.

PknB inhibitors prevent mycobacterial growth.We next looked for strong PknB inhibitors and tested their effect on mycobacterial growth. Since the physiological substrate of PknB is unknown, the initial kinase assays to identify PknB inhibitors (within a panel of 18 commercially available compounds known to inactivate different eukaryotic STPKs) used MBP as a surrogate substrate (4, 9). As shown in Fig. 3a, significant inhibitory effects were observed, in particular for K-252-a and K-252-b, two natural products that contain the indole carbazole chromophore and are thought to target the ATP-binding site. It should be noted that, although the PKA inhibitor H-7 has been previously reported to inhibit PknB (16), we observed no effect of this compound on PknB, whereas it was able to inactivate mouse PKA at a 5 μM concentration in the present assay (data not shown).

• Open in new tab
• Download powerpoint

FIG. 3.

PknB inhibition assays. (a) Autoradiography of a sodium dodecyl sulfate-polyacrylamide electrophoresis gel showing PknB phosphorylation of MBP and the effect of different inhibitors at 100 μM. (b) Autoradiography of labeled spots on phosphocellulose paper showing the effect of the same inhibitors at 10 μM using a more sensitive phosphorylation assay (PknB phosphorylation of GarA). (a and b) Tested inhibitors were: lane 1, A-3; 2, calphostin C; 3, GF109203X; 4, H-7; 5, H-8; 6, H-9; 7, H-89; 8, HA-1004; 9, HA-1077; 10, hypericin; 11, K-252-a; 12, K-252-b; 13, KN-62; 14, KT-5720; 15, KT-5823; 16, ML-7; 17, ML-9; and 18, staurosporine. Positive (+) and negative (−) controls are shown. (c) Autoradiography of labeled spots on phosphocellulose paper showing GarA phosphorylation after serial twofold dilutions, starting at left from 10 μM (K-252-a and K-252-b) or 20 μM (staurosporine). (d) Dose-response plot of PknB fractional activity from five independent experiments as a function of K-252-a concentration. Similar plots were obtained for K-252-b and staurosporine. Vi and Vo, velocities in the presence and absence of inhibitors, respectively.

Although PknB phosphorylates MBP in at least five different sites (17), MBP is a poor PknB substrate; a high enzyme:substrate molar ratio (1:20), which is probably unrepresentative of the physiological situation, has to be used in the kinase assay. We therefore used a proteomic approach to identify putative physiological substrates of PknB and identified the protein GarA, an FHA domain-containing protein, as the optimal PknB substrate in a soluble protein extract from M. tuberculosis (43). Based on the apparent kinetic constants, a new kinase assay was established using a PknB:GarA molar ratio of 1:2,000 and an ATP concentration of 2.25 μM, the apparent Km value for ATP. A screening of the panel of inhibitors using this assay confirmed the previous results, showing K-252-a, K-252-b, and staurosporine to be the strongest inhibitors (Fig. 3b). The IC₅₀ values for these compounds were determined by quantification and graphical analysis of radiolabeled spots from serial twofold dilutions (Fig. 3c) and found to be 96 ± 7 nM for K-252-a (Fig. 3d), 106 ± 6 nM for K-252-b, and 0.6 ± 0.05 μM for staurosporine.

Using a colorimetric, resazurin microtiter assay (28), the antibacterial activity of compounds displaying a strong inhibitory effect in vitro was then assayed against M. tuberculosis H37Rv, M. smegmatis mc²155, and Mycobacterium aurum A+. The latter species is generally very sensitive to antitubercular drugs. The results of the susceptibility tests (Table 1) demonstrate that K-252-a inhibited the growth of both M. tuberculosis H37Rv and M. smegmatis mc²155 at a concentration of 20 μM and that of M. aurum A+ at 5 μM. As a control, staurosporine also showed inhibitory effects on M. tuberculosis H37Rv (25 μM < MIC < 50 μM). In contrast, K-252-b failed to inhibit the growth of all mycobacterial species at the highest concentration tested (40 μM), perhaps due to the low permeability of the envelope to this compound.

To further assess the correlation between PknB inactivation and the inhibition of mycobacterial growth, the M. tuberculosis pknB gene and flanking regions were cloned into pOMK, a mycobacterial replicative vector. The resulting plasmid, pOMK-pknB, was introduced into M. smegmatis mc²155, and the overexpression of PknB was assessed by Western blotting (data not shown). As a control, M. smegmatis mc²155 was also transformed with the empty pOMK vector, and the K-252-a resistance of the mycobacterial transformants was then assayed as described above. Two independent experiments showed that the MICs of the M. smegmatis mc²155 strains transformed with pOMK-pknB were twofold that of the control strain (Table 1), further suggesting that PknB is the actual molecular target in vivo.